Palmitoylethanolamide and Hemp Oil Extract Exert Synergistic Anti-Nociceptive Effects in Mouse Models of Acute and Chronic Pain

Pharmacological Research 167 (2021) 105545

Contents lists available at ScienceDirect

Pharmacological Research

journal homepage: www.elsevier.com/locate/yphrs

Palmitoylethanolamide and hemp oil extract exert synergistic anti-nociceptive effects in mouse models of acute and chronic pain

Alex Mabou Tagne a,1, Yannick Fotio a,1, Lin Lin a, Erica Squire a, Faizy Ahmed a, Tarif Ibne Rashid a, Elnaz Karimian Azari b, Daniele Piomelli a,c,d,* a Department of Anatomy and Neurobiology, University of California, Irvine, CA 92697-4625, USA b Metagenics, Aliso Viejo, CA 92656, USA c Department of Biological Chemistry, University of California, Irvine, CA 92697-4625, USA d Department of Pharmaceutical Sciences, University of California, Irvine, CA 92697-4625, USA

ARTICLE INFO ABSTRACT

Keywords: The use of products derived from hemp – i.e., cannabis varieties with low Δ9-tetrahydrocannabinol (Δ9-THC) Palmitoylethanolamide content – as self-medication for pain and other health conditions is gaining in popularity but preclinical and Hemp clinical evidence for their effectiveness remains very limited. In the present study, we assessed the efficacy of a Acute pain full-spectrum hemp oil extract (HOE; 10, 50 and 100 mg-kg-1; oral route), alone or in combination with the anti- Chronic pain inflammatory and analgesic agent palmitoylethanolamide (PEA; 10, 30, 100 and 300 mg-kg-1; oral route), in the formalin and chronic constriction injury (CCI) tests. We found that HOE exerts modest antinociceptive effects when administered alone, whereas the combination of sub-effective oral doses of HOE and PEA produces a substantial greater-than-additive alleviation of pain-related behaviors. Transcription of interleukin (IL) 6 and IL-10 increased significantly in lumbar spinal cord tissue on day 7 after CCI surgery, an effect that was attenuated to the same extent by HOE alone or by the HOE/PEA combination. Pharmacokinetic experiments show that co- administration of HOE enhances and prolongs systemic exposure to PEA. Collectively, our studies lend support to possible beneficial effects of using HOE in combination with PEA to treat acute and chronic pain.

1. Introduction (Δ9-THC) as the compound that most likely underpins the analgesic properties of cannabis [46]. Δ9-THC and its synthetic mimics exert a Current pain management strategies rely on three main classes of broad spectrum of antinociceptive effects in animal models by engaging drugs: opioids, gabapentinoids and non-steroidal anti-inflammatory G protein-coupled CB1 and CB2 cannabinoid receptors [42]. CB1 reagents. These medications are widely used in clinical practice but pro- ceptors are highly expressed in pain-processing structures of the pe- duce adequate pain relief only in a fraction of patients who take them. In ripheral and central nervous systems as well in non-neural cells addition, their use is associated with dose-limiting side effects, such as throughout the body, while CB2 receptors are primarily concentrated in tolerance, abuse liability and gastrointestinal toxicity. Thus, despite cells of the innate and adaptive immune systems [25]. The presence of important progress in charting the mechanisms that underpin patho- CB1 in brain areas involved in the control of mood, reward and cognition logical pain, there is still an unmet need for safe and effective analgesic underlies the psychotropic properties of Δ9-THC, which limit the use of therapies [44]. this agent in pain therapy [32]. A recent survey by the National Academies of Sciences, Engineering In addition to Δ9-THC, cannabis contains more than 140 phyto- and Medicine (NASEM) concluded that substantial clinical evidence chemicals that share a similar terpenophenolic nucleus but do not pro- indicates that cannabis and cannabinoids are effective in the treatment ductively interact with either CB1 or CB2 receptors. There is evidence of chronic pain in adults [30]. This statement is supported by a large that one such compound, cannabidiol (CBD), might reduce pain without number of preclinical studies which point to Δ9-tetrahydrocannabinol causing intoxication but its analgesic efficacy and mechanism remain

* Correspondence to: Louise Turner Arnold Chair in the Neurosciences Distinguished Professor, Departments of Anatomy and Neurobiology, Pharmaceutical Sciences and Biological Chemistry Director, Center for the Study of Cannabis University of California, Irvine 92697-1275, USA. E-mail address: [email protected] (D. Piomelli). 1 Equal contribution https://doi.org/10.1016/j.phrs.2021.105545 Received 21 January 2021; Received in revised form 8 March 2021; Accepted 9 March 2021 Available online 12 March 2021 1043-6618/© 2021 Elsevier Ltd. All rights reserved. A. Mabou Tagne et al. Pharmacological Research 167 (2021) 105545 unclear [9,11]. Cannabis contains other non-psychotropic cannabinoid Table 1 and terpene molecules, whose contribution to the analgesic properties of Chemical composition of hemp oil extract. Nd-Not detected = less the plant may be far from negligible [17,37]. Indeed, cannabis varieties than 0.01% g/g. 9 that contain less than 0.3% Δ -THC – conventionally referred to as Phytochemical % g/g ’ – ‘hemp are becoming increasingly popular as self-medication for pain Cannabidivarin (CBDV) nd and other health conditions, and the enactment of the U.S. 2018 Farm Cannabidiolic acid (CBDA) 1.43 Bill, which effectively legalized the cultivation and sale of hemp, has Cannabigerol (CBG) nd only increased this popularity. However, we still know surprisingly little Cannabigerolic acid (CBGA) 1.23 about the effectiveness of hemp-based products on pain. Cannabidiol (CBD) 9.34 Tetrahydrocannabivarin (THCV) nd Palmitoylethanolamide (PEA) belongs to the fatty acid ethanolamide Delta-9-tetrahydrocannabinol (THC) nd (also known as N-fatty acylethanolamine) class of signaling lipids, which Cannabichromene (CBC) nd also includes the endocannabinoid anandamide and the satiety factor Tetrahydrocannabinolic acid (THCA) nd oleoylethanolamide [33]. A large body of animal and human studies Cannabichromenic acid (CBCA) nd Cannabielsoin (CBE) 0.04 indicate that PEA produces safe and effective analgesia [16,19,23,31,6], Saturated fat 9.3 ’ which depends on the compound s ability to engage the nuclear receptor ω-9 fatty acids 8.6 peroxisome proliferator-activated receptor-α (PPAR-α) [23,24]. ω-3 fatty acids 21.1 In the present study, we investigated the efficacy of a full-spectrum ω-6 fatty acids 59.9 hemp oil extract (HOE), either alone or in combination with PEA, in two mouse models of acute and chronic pain: the formalin and the chronic-constriction injury tests. The results show that HOE exerts modest antinociceptive effects when administered alone, whereas the Table 2 combination of HOE with PEA results in a substantial alleviation of pain- Concentrations of cannabidiol (CBD) and cannabidiolic acid (CBDA) in HOE related behaviors. Pharmacokinetic studies show that this unexpected over time. Data are expressed as mean ± RSD of 4 serial dilutions, each dilution synergistic interaction may be accounted for, at least in part, by the run in triplicates. ability of HOE to prolong the exposure and lifetime of PEA in circulation. Experiments CBD %, g/g CBDA %, g/g

1. 2.98 ± 1.13 0.59 ± 0.64 2. Materials and methods 2. 2.97 ± 1.92 0.59 ± 1.40 3. 3.33 ± 0.84 0.56 ± 2.93 2.1. Animals

We used male CD-1 mice (9–10 weeks old; Charles River, Wilming- 1 h before formalin injection. Following injection, mice were immedi- ton, MA, US) weighing 25–30 g upon arrival. They were randomly ately transferred to a transparent observation chamber where noci- assigned to treatment groups and housed in ventilated plastic cages (4–5 fensive behavior (time spent licking or biting the injected paw, number per cage) in the animal facility of the University of California, Irvine. of shakings) was continuously recorded by a video camera for 60 min Animals were maintained in a pathogen-free environment (12-h light/ and then measured by trained observers blinded to experimental con- ◦ dark cycle) under controlled temperature (20 ± 2 C) and humidity ditions. Behavioral tests and paw edema measurements were performed (55–60%) with food and water available ad libitum. Mice were allowed on post-formalin day (PFD) 7 in both the injected (ipsilateral) and to acclimate for at least 7 days and experiments were conducted during contralateral paws. the light phase of the light/dark cycle. All efforts were made to minimize the number of animals used and their discomfort. All procedures were 2.4. Chronic constriction injury (CCI) approved by the Institutional Animal Care and Use Committee at the University of California, Irvine, and were carried out in strict accordance We elicited peripheral neuropathy in the left sciatic nerve of mice as with the National Institutes of Health guidelines for the care and use of described elsewhere [4]. In brief, mice were anesthetized in 2–3% iso- animals. flurane in O2. Under aseptic conditions, the sciatic nerve was exposed at mid-thigh level through a small incision and loosely tied at 3 distinct – 2.2. Test compounds preparation and administration sites (spaced at 1-mm interval) with 4 0 chromic catgut (Ethicon, USA). The wound was closed with a single muscle suture (Mersilk 5.0, Ethicon, ◦ HOE (code: VOHO-MGC) and PEA (batch n 20190604_03) were USA) and glue to fasten the skin. Operated mice were returned to their supplied by Metagenics (https://www.metagenics.com) at no charge. home cages for recovery. Test compounds or vehicle were given by the The chemical composition of HOE is presented in Table 1 and the cer- oral route on day 7 after surgery (acute treatment) and then repeatedly tificate of analysis is shown asSupporting information. The extract was once daily for 7 consecutive days (subchronic treatment). Behavioral ◦ aliquoted upon arrival at our facility and kept at 80 C until use. Before tests were performed within 1 h of acute treatment or last subchronic tests, aliquots were thawed at room temperature and the extract was treatment in both operated (ipsilateral) and contralateral paws. Mice diluted to the desired dose in a vehicle of 85% distilled water/15% were sacrificed by decapitation, the spinal cords were extruded by hy- Tween 80. Defrosted samples were discarded after use. HOE was sub- draulic pressure and processed for specific mRNA quantification in jected to an internal quality control to monitor its content of CBD and lumbar segments by real-time quantitative PCR. CBDA over time. The results are shown in Table 2. HOE, PEA and their combinations were administered in volumes of 10 mL/kg by oral 2.5. Behavioral testing gavage. Mechanical sensitivity was assessed using a dynamic plantar aes- ◦ 2.3. Formalin test thesiometer (Cat. N 37450, Ugo Basile, Italy). Mice were placed individually in transparent cages positioned on a wire mesh surface. After a The formalin test was described in detail previously [13]. We 45-min habituation period, a mechanical stimulus was delivered to the injected formalin (1% v/v, 10 µL) or vehicle (saline, 10 µL) into the plantar surface of the hind paws through the metal grid by an automated plantar surface of the right hind paw of male mice. Test compounds or steel filament exerting an increasing force ranging from 0 to 5 g over 10 vehicle (85% distilled water/15% Tween 80) were administered orally s. The force at which the mouse withdrawn its paw (withdrawal

2 A. Mabou Tagne et al. Pharmacological Research 167 (2021) 105545 threshold, in grams) was measured automatically. Thermal sensitivity 2.8. Sample preparation was measured using a Hargreaves plantar test apparatus (San Diego Instruments, Inc, CA, USA). Mice were individually placed in small en- 2.8.1. Extraction of PEA, anandamide (AEA) and 2-arachidonoyl-sn- closures with a glass floor. After a 45-min habituation period, the plantar glycerol (2-AG) surface of the hind paws was exposed to a beam of radiant heat (infrared Extraction method was described elsewhere [2,22]. Plasma (50 µL) heat intensity: 3.0) through the glass floor. The cut-off time was set at 15 was transferred into 8-mL glass vials and diluted with water (0.9 mL) 2 s. The time taken to withdraw the paw from heat stimulus (withdrawal and 50 µL of the following internal standards (ISTD): [ H4]-PEA, 2 2 latency, in seconds) was measured automatically. Each paw was tested 3 [ H4]-AEA, and [ H5] 2-AG (100 nM each). Samples were loaded onto times with a 2-min interval between stimuli and the mean paw with- preconditioned Oasis HLB cartridges (Waters Corporation, Milford drawal threshold and latency were calculated. Paw thickness was Massachusetts, USA) washed with methanol (100%) and water and measured in both ipsilateral and contralateral paws using a digital eluted under vacuum (3–5 mmHg). The cartridges were rinsed with ◦ caliper (Cat. N 06–664–16, Fisher scientific, USA). Paw edema was methanol (40%). Acetonitrile (1 mL) was added, and vacuum pressure quantified as the difference (Δpaw thickness, in mm) between the ipsi- was increased gradually to 10 mmHg to ensure maximal analyte re- lateral paw thickness and the contralateral counterpart. covery. Eluates were dried under nitrogen and reconstituted in 100 µL of acetonitrile. Samples were transferred to deactivated glass inserts (200 µL) and placed inside amber glass vials (2 mL; Agilent Technologies, 2.6. Real-time quantitative PCR Wilmington, DE). Lumbar spinal cords (~15 mg) were transferred into 2 mL Precellys soft tissue tubes (Bertin Instruments, France) and spiked 2 2 2 We extracted total RNA from lumbar spinal cords (L3-L6) using with 50 µL of ISTD ([ H4]-PEA, [ H4]-AEA, and [ H5] 2-AG,100 nM ◦ TRIzol™ reagent (Thermo Fisher Scientific, Walthman, USA) and puri- each) and ice-cold acetone (1 mL). Samples were homogenized at 4 C, fied with the PureLink™ RNA Mini Kit (Invitrogen, Carlsbad, USA) as 6000 rpm, 15 s/cycle for 2 cycles with 20 s pause in between. The su- directed by the supplier. Prior to purification, samples were rendered pernatants were carefully transferred into 8-mL glass vials and dried genomic DNA-free by passing the isolated RNA extract through a gDNA under nitrogen. 3 mL of chloroform/methanol (2:1, v/v) and 1 mL of Eliminator spin column (Qiagen, Germantown, USA). RNA concentra- water were added to the samples, which were then stirred vigorously ◦ tion and purity were determined using the NanoDrop 2000/2000-c and centrifuged at 3000 rpm for 15 min at 4 C. The lower phases were spectrophotometer (Thermo Fisher Scientific, Walthman, USA). cDNA collected, while upper phases were extracted again with chloroform (2 was synthesized using 2 µg of total RNA as input for the High-Capacity mL). Eluates were dried under nitrogen and reconstituted in 100 µL of cDNA RT Kit with RNase Inhibitor (Applied BioSystems, Foster City, acetonitrile. Samples were transferred to deactivated glass inserts (200 USA) with a final reaction volume of 20 µL. First-strand cDNA was µL) and placed inside amber glass vials (2 mL). amplified using TaqMan™ Universal PCR Master Mix (Thermo Fisher Scientific, Walthman, USA) following the manufacturer’s instructions. 2.8.2. CBD and CBDA extraction Real-time PCR primers and fluorogenic probes were purchased from Extraction method was described previously [45]. We weighed HOE Applied Biosystems (TaqMan(R) Gene Expression Assays, Foster City, (8–10 mg) in 8-mL glass vials, added 5 mL of ethanol and stirred CA). We used TaqMan gene expression assays for mouse Actin-β vigorously. The samples were diluted 100-fold with methanol and the 2 (Mm00607939_s1), Hprt (Mm00446968_m1), Gapdh diluted solutions (0.5 mL) were spiked with 5 µL of [ H3]-CBD (5.0 (Mm99999915_g1), Tnf-α (Mm00443258_m1), Il-1β µg/mL) and filtered through a 0.45 µm syringe filter (Agilent Technol- (Mm00434228_m1), Il-6 (Mm00446190_m1), and Il-10 ogies, Wilmington, DE) followed by serial dilutions in methanol directly (Mm00439614_m1) (Applied Biosystems, Foster City, CA). Real-time into 200 µL glass inserts placed in 2.0 mL vials. Plasma (100 µL) was 2 PCR reactions were performed in 96-well plates using CFX96™ Real- transferred into 8-mL glass vials and mixed with 50 µL of [ H3]-CBD Time System (Bio-Rad, USA). The thermal cycling conditions were as (100 ng/mL). Proteins were precipitated by addition of 0.5 mL ice-cold ◦ follows: initial denaturation set at 95 C for 10 min, followed by 45 acetonitrile containing 1% formic acid. Lumbar spinal cord samples ◦ ◦ cycles, where each cycle was performed at 95 C for 30 s and at 55 C for (~15 mg) were transferred into 2 mL Precellys soft tissue tubes and 2 60 s. The Bestkeeper software was used to determine the expression spiked with 50 µL of [ H3]-CBD (100 ng/mL) and 1 mL of ice-cold stability and the geometric mean of three different housekeeping genes acetonitrile containing 1% formic acid. Spinal cord samples were then (Actb, Hprt and Gapdh). ΔCt values were calculated as the difference homogenized as described above, and the supernatants were carefully between the Ct value of the geometric mean of these housekeeping genes transferred into 8-mL glass vials. Plasma and spinal cord supernatants ◦ and the Ct value of the genes of interest. The relative quantity of genes of were stirred vigorously and centrifuged at 3000 rpm for 15 min at 4 C. ΔΔ interest was calculated by the 2- Ct method and expressed as fold After centrifugation, the supernatants were loaded onto 1 mL Captiva change over vehicle control. EMR lipid cartridges (Agilent Technologies, USA) and eluted under vacuum (3–5 mmHg). Tissue pellets were rinsed with water/acetonitrile (1:4, v/v; 0.2 mL), stirred for 30 s, and centrifuged at 3000 rpm for 15 ◦ 2.7. Pharmacokinetic experiments min at 4 C. The supernatants were collected, transferred onto EMR cartridges, eluted, and pooled with the first eluate. The cartridges were We administered HOE (100 mg-kg-1), PEA (10 or 20 mg-kg-1) or the rinsed with water/acetonitrile (1:4, v/v; 0.2 mL). Eluates were dried combinations of both to adult mice in volumes of 10 mL-kg-1 by the oral under a nitrogen and reconstituted in 100 µL of methanol containing route. The animals were deeply anesthetized with isoflurane at various 0.1% formic acid. Samples were transferred to deactivated glass inserts time points following test compounds administration (0, 15, 30, 45 and (200 µL) and placed inside amber glass vials (2 mL). 60 min), blood was collected by cardiac puncture into ethylenediamine- tetraacetic acid (EDTA)-rinsed syringes and transferred into 1 mL 2.9. Liquid chromatography/tandem mass spectrometry (LC-MS/MS) polypropylene plastic tubes containing spray-coated potassium-EDTA analysis × ◦ (K2-EDTA). Plasma was prepared by centrifugation at 1450 g at 4 C for 15 min, and transferred into polypropylene tubes, which were LC separations were carried out using a 1260 series LC system ◦ immediately frozen and stored at 80 C. Animals were euthanized by (Agilent Technologies, Santa Clara, CA), consisting of a binary pump decapitation, their spinal cords were quickly extruded by hydraulic with degasser, thermostated autosampler and column compartment pressure on an ice-cold glass plate and lumbar segments were harvested, coupled to a 6460 C triple-quadrupole mass spectrometric detector ◦ frozen on dry ice and stored 80 C until analyses. (MSD; Agilent Technologies, Santa Clara, CA). Analytes were separated

3 A. Mabou Tagne et al. Pharmacological Research 167 (2021) 105545 on an Eclipse XDB C18 column (1.8 µm, 2.1 × 30.0 or 50.0 mm; Agilent pressure was set at 45–50 psi. The MassHunter software (Agilent Tech- Technologies, Wilmington, DE). For CBD and CBDA analyses, the mobile nologies, Santa Clara, CA) was used for instrument control, data phase consisted of water containing 0.1% formic acid as solvent A and acquisition and analysis. methanol containing 0.1% formic acid as solvent B [45]. Representative LC-MS/MS tracings are illustrated in Supplementary Fig. 1–3. For PEA, 2.10. Statistical analysis AEA and 2-AG analyses, the mobile phase consisted of water containing 0.25% acetic acid and 5 mM ammonium acetate as solvent A and ± Results are presented as mean S.E.M. of n experiments. ED50 values methanol containing 0.25% acetic acid and 5 mM ammonium acetate as were determined by linear regression analysis of dose-response curves. solvent B [2]. The flow rate was 0.3–0.5 mL/min. The step-gradient Individual slopes of the dose-response curves were compared by Stu- conditions were as follows. CBD and CBDA: starting at 72% B for dent’s t-test, according to the test of parallelism. Analyses were con- 1.50 min, changed to 95% B at 1.51 min, and maintained till 2.5 min to ducted using Prism software (GraphPad Software, San Diego, CA, remove any strongly retained materials from the column; the column Version 8.4.2). Areas under the time-course curves (AUC) were calcu- was re-equilibrated for 3.5–5 min to 72% B before the next injection. lated using the trapezoidal rule. Differences between groups were PEA, AEA and 2-AG: starting at 78% B to 28% B in 8.00 min, changed to determined by one- or two-way analysis of variance (ANOVA) followed 95% B at 8.01 min, and maintained till 10.00 min; then changed back to by Dunnett’s test for multiple comparisons, as appropriate. The signifi- 78% B at 10.01 min; the equilibration time was 5 min. The column ◦ ◦ cance level was set at P < 0.05. temperature was maintained at 40 C and the autosampler at 9 C. The total analysis time, including re-equilibration, was 6.0–15.0 min. The 3. Results injection volume was 2.0 µL. To prevent carry over, the needle was washed in the autosampler port for 30 s before each injection, using a 3.1. Antinociceptive effects of HOE wash solution consisting of 10% acetone in water/- methanol/isopropanol/acetonitrile (1:1:1:1, v/v). The MSD was oper- To assess the antinociceptive effects of HOE, we first used the ated in the positive electrospray ionization (ESI) mode, and analytes formalin test as a model of injury-induced spontaneous pain [13]. A were quantified by multiple reaction monitoring (MRM), the acquisition single dose of the extract (10, 50 and 100 mg-kg-1) was given orally to parameters are given in Supplementary Table 1. The capillary and male mice 1 h before intraplantar injection of formalin. As expected nozzle voltages were 3500 V and 300–500 V, respectively. The drying ◦ [13], the chemical irritant produced an immediate nocifensive response gas temperature was 300–350 C with a flow of 9.0–11.0 L/min. Sheath ◦ consisting of two temporally distinct phases of licking and flinching of gas temperature was 300–375 C with a flow of 12 L/min. Nebulizer the injected limb (Fig. 1, A and B; phase I: 0–10 min; phase II: 15–60

Fig. 1. Effects of a single dose of HOE on formalin-evoked acute and persistent nociceptive behaviors. (A) Time-course of the acute nocifensive response to formalin (1%, v/v). (B) Cumulative score of the phase I and phase II of the acute nocifensive response. (C) Paw thickness (injected paw thickness minus non-injected, in mm) at PFD7. (D) Formalin-evoked mechanical allodynia and (E) heat hyperalgesia to both ipsilateral and contralateral paws at PDF7. Data are expressed as mean ± S.E.M (n = 8–10 per group) and analyzed by two-way (A and B) or one-way (C, D and E) ANOVA followed by Dunnett’s test for multiple comparisons. *P < 0.05, **P < 0.01, ***P < 0.001 and ****P < 0.0001 vs. Sham or Vehicle controls.

4 A. Mabou Tagne et al. Pharmacological Research 167 (2021) 105545 min). On post formalin day (PFD) 7, the acute pain event gave way to from 21.8 to 29.1 mg-kg-1; Fig. 4, D and E) by up to 85% (P < 0.0001 vs. edema in the injected paw (Fig. 1, C) and hypersensitivity in both ipsi- vehicle) and 109% (P < 0.0001 vs. vehicle), respectively. lateral and contralateral paws (Fig. 1, D and E). The two highest doses of PEA (100 or 300 mg-kg-1) were adminis- The results show that HOE (100 mg-kg-1) administration had no ef- tered orally to neuropathic mice once daily for 7 consecutive days. The fect on the nocifensive score in Phase I, but reduced Phase II behavior by treatment reversed mechanical allodynia (Fig. 4, C) and heat hyper- 27% (P < 0.05 vs. vehicle) (Fig. 1, A–B). On PDF7, mechanical allodynia algesia (Fig. 4, F), supporting prior studies indicating that there is no was also attenuated by 27% (P < 0.01 vs. vehicle) and 62% (P < 0.01 vs. tolerance to the antinociceptive effects of PEA [24]. vehicle), respectively, in ipsilateral and contralateral paws of mice that -1 had received HOE (100 mg-kg ) (Fig. 1, D). Heat hyperalgesia was 3.3. Antinociceptive effects of combinations of HOE and PEA alleviated by 41% (P < 0.001 vs. vehicle) and 50% (P < 0.001 vs. vehicle) in ipsilateral and contralateral paws, respectively (Fig. 1, E). Next, we tested whether HOE and PEA might interact when given in -1 < Moreover, HOE reduced paw edema by 63% at 50 mg-kg (P 0.01 vs. combination. Our results did not allow us to determine a median -1 > vehicle) but had no such effect at 100 mg-kg (P 0.05 vs. vehicle) effective dose for HOE in either of the two experimental pain models, (Fig. 1, C). which made it impossible to evaluate the pharmacodynamic interactions Next, we produced peripheral neuropathy in mice by ligating their between PEA and HOE by isobolographic analysis. Instead, we opted for right sciatic nerve, which resulted in the development of mechanical a combination subthresholding approach which is a valid, effect-based allodynia and heat hyperalgesia in the operated limb (Fig. 2). On day 7 strategy to investigate the biological activity of combinations of two after surgery, we administered HOE by the oral route 1 h before testing. or more compounds [14]. We administered combinations of The results show that administration of a single dose of HOE (100 mg- sub-effective oral doses of HOE (50 mg-kg-1) and PEA (30 or -1 kg ) had no effect on either heat hyperalgesia or mechanical allodynia 100 mg-kg-1) 1 h before formalin (1% v/v). The results show that the > (Fig. 2, A and B; P 0.05 vs. vehicle), whereas a 7-day treatment with HOE (50 mg-kg-1) and PEA (30 mg-kg-1) combination did not affect -1 the extract (100 mg-kg ) decreased heat hyperalgesia by 33% (Fig. 2, C Phase I of formalin-evoked nocifensive behavior but reduced Phase II by < and D; P 0.01 vs vehicle). 32% (P < 0.01 vs. vehicle) (Fig. 5, A and B). There was no effect on paw edema (Fig. 5, C). 3.2. Antinociceptive effects of PEA On PFD7, formalin-evoked mechanical allodynia was attenuated by 49% (P < 0.0001 vs. vehicle) and 75% (P < 0.0001 vs. vehicle) As expected from previous work [6,24], administration of PEA (10, respectively, in the ipsilateral and contralateral paws of mice that had -1 -1 30, 100 and 300 mg-kg-1, oral) resulted in a dose-dependent suppression received HOE (50 mg-kg ) plus PEA (100 mg.kg ) (Fig. 5, D). Likewise, < of formalin-evoked nocifensive (Fig. 3, A–B) and inflammatory Fig.( 3C) heat hyperalgesia was decreased by 54% (P 0.0001 vs. vehicle) and < responses. Mechanical allodynia on PFD7 was also attenuated by 91% 57% (P 0.0001 vs. vehicle), respectively, in the ipsilateral and (P < 0.0001 vs. vehicle) and 112% (P < 0.0001 vs. vehicle), respec- contralateral paws (Fig. 5, E). tively in ipsilateral and contralateral paws of mice that had received the Finally, we asked whether the combination of HOE and PEA might highest dose of PEA (300 mg-kg-1) (Fig. 3, D). A similar effect was seen affect hypersensitivity in the CCI model. We administered combinations -1 with heat hyperalgesia, which was alleviated by 93% (P < 0.0001 vs. of sub-effective oral doses of HOE (100 mg-kg ) and PEA (10 or 20 mg- -1 vehicle) and 110% (P < 0.0001 vs. vehicle) in ipsilateral and contra- kg ). Fig. 6 shows that the combinations reduced mechanical allodynia lateral paws, respectively (Fig. 3, E). At dosages lower than 300 mg-kg-1, (Fig. 6, A) and heat hyperalgesia (Fig. 6, B) in the operated limbs by as < < PEA had no effect on either mechanical allodynia or heat hyperalgesia much as 85% (P 0.001 vs. vehicle) and 118% (P 0.001 vs. vehicle), (P > 0.05 vs. vehicle). respectively. In the CCI model, a single PEA administration produced strong dose- = -1 dependent suppression of mechanical allodynia (ED50 22.2 mg-kg ; 3.4. RT-qPCR measurements 95% confidence interval [CI] ranging from 18.7 to 26.0 mg-kg-1; Fig. 4, = -1 -1 -1 A and B) and heat hyperalgesia (ED50 25.1 mg-kg ; 95% CI ranging We examined whether HOE (100 mg-kg ), PEA (20 mg-kg ) or their

Fig. 2. Effects of HOE on CCI-evoked mechanical allodynia (A and C) and heat hyperalgesia (B and D). A single dose of HOE (100 mg.kg-1) was administered orally to neuropathic mice on day 7 after CCI surgery (acute) and then repeatedly for 7 consecutive days (subchronic). (A and B) Mechanical allodynia and heat hyperalgesia on day 7 after CCI surgery. (C and D) Mechanical allodynia and heat hyperalgesia on day 14 after CCI surgery. Data are expressed as mean ± S.E.M (n = 8 per group) and analyzed by oneway ANOVA followed by Dunnett’s test for multiple comparisons. **P < 0.01 vs. Vehicle (V) and ****P < 0.0001 vs. Contralateral (Ctra; from vehicle-treated mice) controls.

5 A. Mabou Tagne et al. Pharmacological Research 167 (2021) 105545

Fig. 3. Effects of a single dose of PEA on formalin-evoked acute and persistent nociceptive behaviors. (A) Time-course of the acute nocifensive response to 1% formalin. (B) Cumulative score of the phase I and phase II of the acute nocifensive response. (C) Paw thickness (injected paw thickness minus non injected, in mm) at PFD7. (D) Mechanical and (E) heat sensitivities to both ipsilateral and contralateral paws at PDF7. Data are expressed as mean ± S.E.M (n = 8–12 per group) and analyzed by two-way (A and B) or one-way ANOVA (C, D and E) followed by Dunnett’s test for multiple comparisons. *P < 0.05, **P < 0.01, ***P < 0.001 and ****P < 0.0001 vs. Sham or Vehicle controls. combination might alter the transcription of proalgesic and inflamma- bioavailability of PEA and prolongs its lifetime in circulation. tory cytokines in lumbar spinal cords. The results show that CCI was To assess whether PEA might affect the pharmacokinetic profile of associated with statistically detectable increases in IL-6 and IL-10 mRNA HOE, we administered HOE, either alone or in combination with PEA, expression (Fig. 7, A and B). IL-6 increased 2 folds (P < 0.05 vs. sham) and measured plasma concentrations of two main components of HOE, while IL-10 increased 2.5 folds (P < 0.01 vs. sham). IL-1β mRNA levels CBD and CBDA. The results show that plasma CBD levels increased were also increased but this trend failed to reach statistical significance rapidly to 1676 ± 289.6 pmol-mL-1 (P < 0.05 vs. baseline) and (Fig. 7, C). By contrast, there were no statistically detectable changes in 1826 ± 635.6 pmol-mL-1 (P < 0.05 vs. baseline) after administration of the expression of TNF-α mRNA (Fig. 7, D). Oral administration of a HOE or the combination of HOE with PEA, respectively (Fig. 8, B). single dose of HOE (100 mg-kg-1) to neuropathic mice normalized IL-6 Likewise, CBDA increased up to 39,917 ± 2042 pmol-mL-1 and (P < 0.01 vs. vehicle) and IL-10 (P < 0.01 vs. vehicle) mRNA levels 38,971 ± 5289 pmol-mL-1 after administration of HOE or the combi- (Fig. 7, A and B). When administered alone, PEA (20 mg-kg-1) halved IL- nation of HOE with PEA, respectively (Fig. 8, C). However, the combi- 10 mRNA levels (P < 0.05 vs. vehicle). The combination of HOE nation of HOE with PEA [AUC (CBD) = 52,153 ± 13,999 pmol-min-mL- (100 mg-kg-1) and PEA (20 mg-kg-1) caused a 1.7-fold decrease in IL-6 1; AUC (CBDA) = 142,1014 ± 162,688 pmol-min-mL-1] did not alter the mRNA expression (P < 0.05 vs. vehicle), while lowering IL-10 mRNA overall exposure to CBD (Fig. 8, B) and CBDA (Fig. 8, C) compared to levels 3.7 folds (P < 0.001 vs. vehicle). HOE alone [AUC (CBD) = 56,334 ± 7757 pmol-min-mL-1; AUC (CBDA) = 14,15702 ± 70,988 pmol-min-mL-1], suggesting that PEA 3.5. Pharmacokinetic profiles of HOE, PEA or their combination does not alter the pharmacokinetic properties of HOE. Fig. 9 shows concentration-time curves and overall exposure to PEA We measured the concentrations of PEA, CBD and CBDA in mouse (A), CBD (B) and CBDA (C) in mouse lumbar spinal cords. Baseline levels ± -1 plasma (Fig. 8) and spinal cord tissue (Fig. 9) in a 60-min period of PEA were 117.1 6.3 pmol-mg of tissue (Fig. 9, A). Oral adminis- following oral administration of a single dose of the compounds or their tration of HOE, PEA or their combination had no statistically detectable combination. Fig. 8 shows concentration-time curves and overall expo- effect on such levels in the 60-min time frame of this study. This finding sure (area-under-the-curve, AUC) to PEA (A), CBD (B) and CBDA (C). suggests that PEA does not readily enter the spinal cord and may exhibit Baseline levels of PEA were 10.03 ± 0.4 pmol-mL-1. Oral administration its analgesic action mainly through a peripheral mechanism [6]. By of HOE did not affect baseline PEA concentrations (Fig. 8, A). By contrast, we were able to detect appreciable amounts of CBD and CBDA contrast, plasma PEA levels increased rapidly by up to 44.1 ± 2.4 pmol- in spinal cord after oral administration of HOE alone or in combination mL-1 (P < 0.01 vs. baseline) and 66.9 ± 4.9 pmol-mL-1 (P < 0.05 vs. with PEA (Fig. 9, B–C). Of note, oral administration of the combination = ± -1 baseline) after administration of PEA or the combination of HOE with of HOE with PEA [AUC (CBD) 29.21 7.36 pmol-min-mg ; AUC = ± -1 PEA, respectively. Importantly, the combination of HOE with PEA (CBDA) 43.28 13.64 pmol-min-mg ] did not statistically change (AUC = 2211 ± 208 pmol-min-mL-1) resulted in greater exposure to the overall exposure of mice to CBD and CBDA compared with HOE -1 = ± -1 = PEA than HOE (AUC = 72.1 ± 83.9 pmol-min-mL ; P < 0.0001 vs alone [AUC (CBD) 36.3 5.78 pmol-min-mg ; AUC (CBDA) -1 ± -1 HOE + PEA) and PEA (AUC = 1516 ± 133.2 pmol-min-mL ; P < 0.05 59.04 17.48 pmol-min-mg ], suggesting that PEA does not affect the vs HOE + PEA) given separately, suggesting that HOE enhances the distribution of hemp in the spinal cord.

6 A. Mabou Tagne et al. Pharmacological Research 167 (2021) 105545

Fig. 4. Effects of PEA on CCI-evoked mechanical allodynia (A-C) and heat hyperalgesia (D-F). PEA was administered orally to neuropathic mice on day 7 after CCI surgery (acute) and then repeatedly for 7 consecutive days (subchronic). Dose-response curves for the effects of PEA on CCI-induced mechanical allodynia (A, B) and heat hyperalgesia (D, E) on day 7 after surgery. Effects of subchronic administration of PEA (100 and 300 mg.kg-1) on CCI-induced mechanical allodynia (C) and heat hyperalgesia (F) on day 14 after surgery. Data are expressed as mean ± S.E.M (n = 8 per group) and analyzed by one-way ANOVA followed by Dunnett’s test for multiple comparisons. **P < 0.01 vs. Vehicle (V) and ****P < 0.0001 vs. Contralateral (Ctra) controls. %MPE, % maximal possible effect.

3.6. Effects of HOE alone or in combination with PEA on greater-than-additive alleviation of pain-related behaviors. Pharmaco- endocannabinoid levels kinetic studies suggest that one possible explanation for this synergistic interaction lies in the ability of HOE to prolong the exposure and lifetime We measured the concentrations of the endocannabinoids AEA and of PEA in circulation. 2-AG in mouse plasma and spinal cord tissue after oral administration of Hemp contains a wide range of phytocannabinoids, terpenes, flavo- a single dose of HOE (100 mg-kg-1) alone or in combination with PEA noids and other bioactive molecules which likely contribute to its bio- (20 mg-kg-1). Fig. 10 shows that HOE alone or in combination with PEA logical effects both individually and synergistically [36]. Identifying did not affect plasma (Fig. 10, A and B) or spinal cord (Fig. 10, C and D) which of these myriad chemicals might be responsible for the modest levels of the compounds (P > 0.05 vs baseline) in the 60-min time frame antinociceptive properties of HOE would be a daunting task, but of this study. These findings suggest that acute administration of HOE promising candidates include, among others, CBD and CBDA. For alone or in combination with PEA does not affect the endocannabinoid example, it has been shown that CBD produces antinociceptive effects in levels in mice. some animal models without causing apparent tolerance [18,20,26,47, 49,5,8,9]. Likewise, studies have shown that CBDA exhibits anti- 4. Discussion nociceptive and anti-inflammatory properties in rodents [35,50]. Cellular components of the immune system are thought to contribute Legislative changes have increased the availability of products to pain initiation and maintenance through the release of multiple cy- derived from industrial hemp (i.e., cannabis containing < 0.3% Δ9- tokines [34,40,43]. The present results show that mRNA levels of IL-6 THC), which are now widely used for self-medication of pain and other and IL-10 increased significantly in lumbar spinal cord tissue on day 7 medical conditions [7]. Still very little is known, however, about the after CCI surgery, an effect that was reduced by HOE alone or by the pharmacological properties of such products. In the present study, we HOE/PEA combination. The increase in spinal IL-6 and IL-10 tran- set out to assess the efficacy of a full spectrum HOE, alone or in com- scription is consistent with previous reports [12,15,3,39]. The finding bination with PEA, in two complementary mouse models of pain, the that CCI had little to no impact on IL-1β and TNF-α mRNAs is also in line formalin and CCI tests. We found that HOE has little or no effect when with prior studies [10,39,41]. Oral administration of HOE either alone, administered alone but synergizes with PEA to produce a or in combination with PEA caused a reduction in the mRNA expression

7 A. Mabou Tagne et al. Pharmacological Research 167 (2021) 105545

Fig. 5. Effects of combinations of HOE with PEA on formalin-evoked acute and persistent nociceptive behaviors. (A) Time-course of the acute nocifensive response to 1% formalin. (B) Cumulative score of the phase I and phase II of the acute nocifensive response. (C) Paw thickness (injected paw thickness minus non injected, in mm) at PFD7. (D) Mechanical and (E) heat sensitivities to both ipsilateral and contralateral paws at PDF7. Data are expressed as mean ± S.E.M (n = 8–10 per group) and analyzed by two-way (A and B) or one-way ANOVA followed by Dunnett’s (C, D and E) test for multiple comparisons. *P < 0.05, **P < 0.01, ***P < 0.001 and ****P < 0.0001. N: naïve; ns: nonsignificant.

Fig. 6. Effects of combinations of HOE with PEA on CCI-evoked mechanical allodynia (A) and heat hyperalgesia (B) to operated limbs. A single dose of HOE (100 mg.kg-1), alone or in combination with PEA (10 mg.kg-1) or PEA (20 mg.kg-1) was administered orally to neuropathic mice 1 h before testing. Data are expressed as mean ± S.E.M. (n = 8 per group) and analyzed by one-way ANOVA followed by Dunnett’s test for multiple comparisons. * P < 0.05, **P < 0.01, ***P < 0.001 and ****P < 0.0001. C, contralateral.

of IL-6 and IL-10. This finding is relevant from a therapeutic perspective cannabinoid receptors, respectively [38]. This mechanism is unlikely to as the inhibition of IL-6 with neutralizing antibodies or other pharma- be operational in our studies, however, because we found no evidence cological tools has been found to alleviate nociceptive behaviors sug- for enhanced endocannabinoid signaling after administration of the gestive of neuropathic pain [1,28]. HOE/PEA combination, at least in plasma and spinal cord tissue at the An important finding of the present report is that one or more time of measurement. While other pharmacodynamic interactions might chemical constituents of hemp interact(s) with the endogenous anal- have occurred, our results point to the possibility that the synergistic gesic and anti-inflammatory compound PEA to produce greater-than- potentiation between HOE and PEA could have a pharmacokinetic un- additive antinociceptive effects. This synergism may be underpinned derpinning. Indeed, we found that administering a combination of HOE by pharmacodynamic and/or pharmacokinetic mechanisms. An anti- and PEA enhances and prolongs the systemic exposure to PEA, nociceptive synergism has been previously demonstrated between PEA compared to administering PEA alone. By contrast, PEA did not affect and the endocannabinoid anandamide, which act via PPAR-α and CB1 the pharmacokinetic properties of two quantitatively major constituents

8 A. Mabou Tagne et al. Pharmacological Research 167 (2021) 105545

Fig. 7. Effects of HOE (100 mg.kg-1), PEA (20 mg.kg-1) or their combination on mRNA levels (AU, Arbitrary Units) of IL-6 (A), IL-10 (B), IL-1b (C) and TNF-α (D) in lumbar spinal cords (L3-L6). Bars represent mean ± S.E.M. (n = 3–4 per group) and analyzed by oneway ANOVA followed by Dunnett’s test for multiple comparisons. *P < 0.05, **P < 0.01 and ***P < 0.001 vs. Vehicle. S, sham; ns, non-significant.

Fig. 8. Concentration-time curves and overall drug exposure (AUC) for PEA (A), CBD (B) and CBDA (C) in mouse plasma, after oral administration of HOE (100 mg. kg-1) alone or in combination with PEA (20 mg.kg-1). Data are expressed as mean ± S.E.M (n = 3–5 per group) and analyzed by one-way ANOVA (AUC curves) and two-way ANOVA (concentration-time curves) followed by Dunnett’s test for multiple comparisons. **P < 0.01 vs. baseline or combination of HOE plus PEA.

9 A. Mabou Tagne et al. Pharmacological Research 167 (2021) 105545

Fig. 9. Concentration-time curves and overall drug exposure (AUC) for PEA (A), CBD (B) and CBDA (C) in lumbar spinal cord tissue, after oral administration of HOE (100 mg.kg-1) alone or in combination with PEA (20 mg.kg-1). Data are expressed as mean ± S.E.M. (n = 3–5 per group) and analyzed by one-way ANOVA (AUC curves) and two-way ANOVA (concentration-time curves) followed by Dunnett’s test for multiple comparisons. *P < 0.05 vs. baseline or combination of HOE plus PEA.

Fig. 10. Time-course of the effects of oral administration of HOE (100 mg.kg-1; •) alone or in combination with PEA (20 mg.kg-1; ￭) on the levels of anandamide (AEA) and 2- arachidonoyl-sn-glycerol (2-AG) in mouse plasma (Upper panels) and spinal cords (lower panels). Data are expressed as mean} S.E.M. (n = 3–5 per group) and analyzed by twoway ANOVA followed by Dunnett’s test for multiple comparisons. of HOE, CBD and CBDA. The mechanism underpinning the pharmaco- concentrations [21]. kinetic interaction between HOE and PEA is unknown, but two specu- This study has several limitations. First, we used combination sub- lative scenarios are worth mentioning. First, HOE could act as a carrier thresholding rather than isobolographic analysis to test the interaction to increase the intestinal absorption of PEA. This is sometimes observed between HOE and PEA. This selection was imposed by the weak effects with plant extracts, especially if they contain terpene compounds [48]. of HOE, which did not allow us to determine a median effective dose for Second, HOE could slow down the liver degradation of PEA, which is the compound. Though combination subthresholding is well estab- primarily mediated by FAAH. Evidence in the literature suggests that lished, it fails to define the doses at which the combination of HOE/PEA CBD, which is present in HOE, inhibits FAAH activity at micromolar can be expected to have synergistic effects, which can only be addressed

10 A. Mabou Tagne et al. Pharmacological Research 167 (2021) 105545 by dose-effect-based approaches [14]. Second, we only investigated female rats with persistent inflammatory pain, J. Pharmacol. Exp. Ther. 373 (3) – male mice. As growing evidence indicates that there are significant (2020) 416 428, https://doi.org/10.1124/jpet.119.263319. [6] A. Calignano, G. La Rana, A. Giuffrida, D. Piomelli, Control of pain initiation by sex-related differences in pain responses [27], future work will need to endogenous cannabinoids, Nature 394 (6690) (1998) 277–281, https://doi.org/ include female animals. To the best of our knowledge, sexual di- 10.1038/28393. morphisms have been observed in the antinociceptive effects of some [7] J. Corroon, J.A. Phillips, A cross-sectional study of cannabidiol users, Cannabis Cannabinoid Res. 3 (1) (2018) 152–161, https://doi.org/10.1089/can.2018.0006. components of hemp, including CBD [5], but not in the response to PEA. [8] B. Costa, G. Giagnoni, C. Franke, A.E. Trovato, M. Colleoni, Vanilloid TRPV1 Third, our pharmacokinetic studies were focused on CBD and CBDA, receptor mediates the antihyperalgesic effect of the nonpsychoactive cannabinoid, even though many other naturally occurring chemical constituents in cannabidiol, in a rat model of acute inflammation, Br. J. Pharmacol. 143 (2) (2004) 247–250, https://doi.org/10.1038/sj.bjp.0705920. HOE may impact the pharmacokinetic properties of PEA. Lastly, the [9] B. Costa, A.E. Trovato, F. Comelli, G. Giagnoni, M. Colleoni, The non-psychoactive chemical composition of hemp and hemp-derived products varies cannabis constituent cannabidiol is an orally effective therapeutic agent in rat – considerably depending on various factors including growth conditions chronic inflammatory and neuropathic pain, Eur. J. Pharmacol. 556 (1 3) (2007) 75–83, https://doi.org/10.1016/j.ejphar.2006.11.006. and extraction methods [29], which may render difficult a head-to-head [10] V. Curto-Reyes, G. Kirschmann, M. Pertin, S.K. Drexler, I. Decosterd, M.R. Suter, efficacy comparison across different products. Neuropathic pain phenotype does not involve the NLRP3 inflammasome and its In conclusion, our findings identify an unexpected synergistic end product interleukin-1β in the Mice spared nerve injury model, PLoS One 10 (7) (2015), e0133707, https://doi.org/10.1371/journal.pone.0133707. interaction between THC-free hemp and the endogenous analgesic and [11] D. De Gregorio, R.J. McLaughlin, L. Posa, R. Ochoa-Sanchez, J. Enns, M. Lopez- anti-inflammatory factor PEA, such that the combination of the two Canul, M. Aboud, S. Maione, S. Comai, G. Gobbi, Cannabidiol modulates compounds exerts greater-than-additive antinociceptive effects in serotonergic transmission and reverses both allodynia and anxiety-like behavior in – mouse models of acute and chronic pain. This synergism may be at least a model of neuropathic pain, Pain 160 (1) (2019) 136 150, https://doi.org/ 10.1097/j.pain.0000000000001386. ´ ´ ´ ´ ˇ ´ partially ascribed to the ability of HOE to increase the systemic exposure [12] P. Dubový, V. Brazda, I. Klusakova, I. Hradilova-Svízenska, Bilateral elevation of to PEA after oral administration. interleukin-6 protein and mRNA in both lumbar and cervical dorsal root ganglia following unilateral chronic compression injury of the sciatic nerve, J. Neuroinflamm. (2013) 10, https://doi.org/10.1186/1742-2094-10-55. Significance statement [13] D. Dubuisson, S.G. Dennis, The formalin test: a quantitative study of the analgesic effects of morphine, meperidine, and brain stem stimulation in rats and cats, Pain 4 – There is an unmet medical need for safe and effective analgesics. Our (1977) 161 174, https://doi.org/10.1016/0304-3959(77)90130-0. [14] J. Foucquier, M. Guedj, Analysis of drug combinations: current methodological results suggest that hemp oil extract is modestly effective, when landscape, in: Pharmacology Research and Perspectives, 3, Wiley-Blackwell administered alone, in treating acute and chronic pain. Surprisingly, Publishing Ltd., 2015, p. 149, https://doi.org/10.1002/prp2.149. however, the extract strongly enhanced the analgesic effects of palmi- [15] S. Franchi, A.E. Valsecchi, E. Borsani, P. Procacci, D. Ferrari, C. Zaffa, P. Sartori, L. F. Rodella, A. Vescovi, S. Maione, F. Rossi, P. Sacerdote, M. Colleoni, A.E. Panerai, toylethanolamide, an endogenous lipid mediator, such that the combi- Intravenous neural stem cells abolish nociceptive hypersensitivity and trigger nation of the two compounds exerted greater-than-additive nerve regeneration in experimental neuropathy, Pain 153 (4) (2012) 850–861, antinociceptive effects in two mouse models of pain. This synergism may https://doi.org/10.1016/j.pain.2012.01.008. [16] L. Gabrielsson, S. Mattsson, C.J. Fowler, Palmitoylethanolamide for the treatment be harnessed therapeutically to produce adequate pain relief without of pain: pharmacokinetics, safety and efficacy, British Journal of Clinical causing undesirable side effects. Pharmacology (2016) 932–942, https://doi.org/10.1111/bcp.13020. ˇ [17] R. Gallily, Z. Yekhtin, L.O. Hanus, The anti-inflammatory properties of terpenoids from cannabis, Cannabis Cannabinoid Res. 3 (1) (2018) 282–290, https://doi.org/ CRediT authorship contribution statement 10.1089/can.2018.0014. [18] K. Genaro, D. Fabris, A.L.F. Arantes, A.W. Zuardi, J.A.S. Crippa, W.A. Prado, Participated in research design: Piomelli, Mabou Tagne, Fotio and Cannabidiol is a potential therapeutic for the affective-motivational dimension of incision pain in rats, Front. Pharmacol. 8 (2017) 391, https://doi.org/10.3389/ Karimian Azari. Conducted experiments: Mabou Tagne, Fotio, Lin, fphar.2017.00391. ´ Squire, Ahmed and Ibne Rashid. Contributed new reagents or analytic [19] J. Guindon, J. LoVerme, A. De Lean, D. Piomelli, P. Beaulieu, Synergistic tools: Karimian Azari. Performed data analysis: Mabou Tagne and Fotio. antinociceptive effects of anandamide, an endocannabinoid, and nonsteroidal anti- inflammatory drugs in peripheral tissue: a role for endogenous fatty-acid Wrote or contributed to the writing of the manuscript: Piomelli, Mabou ethanolamides? Eur. J. Pharmacol. 550 (1–3) (2006) 68–77, https://doi.org/ Tagne and Fotio. 10.1016/j.ejphar.2006.08.045. [20] D.C. Hammell, L.P. Zhang, F. Ma, S.M. Abshire, S.L. McIlwrath, A.L. Stinchcomb, K. N. Westlund, Transdermal cannabidiol reduces inflammation and pain-related Conflict of interest behaviours in a rat model of arthritis, Eur. J. Pain. 20 (6) (2016) 936–948, https:// doi.org/10.1002/ejp.818. [21] F.M. Leweke, D. Piomelli, F. Pahlisch, D. Muhl, C.W. Gerth, C. Hoyer, ¨ Dr. Elnaz Karimian Azari is an employee of Metagenics, which pro- J. Klosterkotter, M. Hellmich, D. Koethe, Cannabidiol enhances anandamide vided financial support for the present study. signaling and alleviates psychotic symptoms of schizophrenia, Transl. Psychiatry 2 (3) (2012), e94, https://doi.org/10.1038/tp.2012.15. [22] L. Lin, H. Yang, P.J.H. Jones, Quantitative analysis of multiple fatty acid Appendix A. Supporting information ethanolamides using ultra-performance liquid chromatography-tandem mass spectrometry, Prostaglandins Leukot. Essent. Fat. Acids 87 (6) (2012) 189–195, Supplementary data associated with this article can be found in the https://doi.org/10.1016/j.plefa.2012.09.001. [23] J. Lo Verme, J. Fu, G. Astarita, G. La Rana, R. Russo, A. Calignano, D. Piomelli, The online version at doi:10.1016/j.phrs.2021.105545. nuclear receptor peroxisome proliferator-activated receptor-α mediates the anti- inflammatory actions of palmitoylethanolamide, Mol. Pharmacol. 67 (1) (2005) References 15–19, https://doi.org/10.1124/mol.104.006353. [24] J. LoVerme, R. Russo, G. La Rana, J. Fu, J. Farthing, G. Mattace-Raso, R. Meli, A. Hohmann, A. Calignano, D. Piomelli, Rapid broad-spectrum analgesia through [1] J.L. Arruda, S. Sweitzer, M.D. Rutkowski, J.A. Deleo, Intrathecal anti-IL-6 antibody activation of peroxisome proliferator-activated receptor-α, J. Pharmacol. Exp. and IgG attenuates peripheral nerve injury-induced mechanical allodynia in the Ther. 319 (3) (2006) 1051–1061, https://doi.org/10.1124/jpet.106.111385. rat: possible immune modulation in neuropathic pain, Brain Res. 879 (1–2) (2000) [25] K. Mackie, Cannabinoid receptors as therapeutic targets, Annu. Rev. Pharmacol. 216–225, https://doi.org/10.1016/S0006-8993(00)02807-9. Toxicol. 46 (1) (2006) 101–122, https://doi.org/10.1146/annurev. [2] G. Astarita, D. Piomelli, Lipidomic analysis of endocannabinoid metabolism in pharmtox.46.120604.141254. biological samples, J. Chromatogr. B Anal. Technol. Biomed. Life Sci. 877 (26) [26] A.M. Malfait, R. Gallily, P.F. Sumariwalla, A.S. Malik, E. Andreakos, (2009) 2755–2767, https://doi.org/10.1016/j.jchromb.2009.01.008. R. Mechoulam, M. Feldmann, The nonpsychoactive cannabis constituent [3] P.J. Austin, G. Moalem-Taylor, The neuro-immune balance in neuropathic pain: cannabidiol is an oral anti-arthritic therapeutic in murine collagen-induced involvement of inflammatory immune cells, immune-like glial cells and cytokines, arthritis, Proc. Natl. Acad. Sci. U.S.A. 97 (17) (2000) 9561–9566, https://doi.org/ J. Neuroimmunol. 229 (2010) 26–50. 10.1073/pnas.160105897. [4] G.J. Bennett, Y.K. Xie, A peripheral mononeuropathy in rat that produces disorders [27] J.S. Mogil, Sex differences in pain and pain inhibition: multiple explanations of a of pain sensation like those seen in man, Pain 33 (1) (1988) 87–107, https://doi. controversial phenomenon, in: Nature Reviews Neuroscience, 13, Nature org/10.1016/0304-3959(88)90209-6. Publishing Group, 2012, pp. 859–866, https://doi.org/10.1038/nrn3360. [5] S.C. Britch, A.G. Goodman, J.L. Wiley, A.M. Pondelick, R.M. Craft, Antinociceptive and immune effects of delta-9-tetrahydrocannabinol or cannabidiol in male versus

11 A. Mabou Tagne et al. Pharmacological Research 167 (2021) 105545

[28] T. Murakami, T. Kanchiku, H. Suzuki, Y. Imajo, Y. Yoshida, H. Nomura, D. Cui, [40] J. Scholz, C.J. Woolf, The neuropathic pain triad: neurons, immune cells and glia, T. Ishikawa, E. Ikeda, T. Taguchi, Anti-interleukin-6 receptor antibody reduces in: Nature Neuroscience, 10, Nature Publishing Group, 2007, pp. 1361–1368, neuropathic pain following spinal cord injury in mice, Exp. Ther. Med. 6 (5) (2013) https://doi.org/10.1038/nn1992. ˜ 1194–1198, https://doi.org/10.3892/etm.2013.1296. [41] K.A.B. Simao da Silva, A.F. Paszcuk, G.F. Passos, E.S. Silva, A.F. Bento, F.C. Meotti, [29] D. Namdar, M. Mazuz, A. Ion, H. Koltai, Variation in the compositions of J.B. Calixto, Activation of cannabinoid receptors by the pentacyclic triterpene cannabinoid and terpenoids in Cannabis sativa derived from inflorescence position α,β-amyrin inhibits inflammatory and neuropathic persistent pain in mice, Pain along the stem and extraction methods, Ind. Crops Prod. 113 (2018) 376–382, 152 (8) (2011) 1872–1887, https://doi.org/10.1016/j.pain.2011.04.005. https://doi.org/10.1016/j.indcrop.2018.01.060. [42] K. Starowicz, D.P. Finn, Cannabinoids and pain: sites and mechanisms of action, in: [30] National Academies of Sciences, E. and M, The health effects of cannabis and Advances in Pharmacology, 80, Academic Press Inc., 2017, pp. 437–475, https:// cannabinoids: the current state of evidence and recommendations for research, in: doi.org/10.1016/bs.apha.2017.05.003. Psychiatria, 15, Via Medica, 2017, https://doi.org/10.17226/24625. [43] M.A. Thacker, A.K. Clark, F. Marchand, S.B. McMahon, Pathophysiology of [31] S. Petrosino, V. Di Marzo, The pharmacology of palmitoylethanolamide and first peripheral neuropathic pain: immune cells and molecules, Anesth. Analg. 105 (3) data on the therapeutic efficacy of some of its new formulations, in: British Journal (2007) 838–847, https://doi.org/10.1213/01.ane.0000275190.42912.37. of Pharmacology, 174, John Wiley and Sons Inc, 2017, pp. 1349–1365, https://doi. [44] G. Varrassi, E. Alon, M. Bagnasco, L. Lanata, V. Mayoral-Rojals, A. Paladini, J. ¨ org/10.1111/bph.13580. V. Pergolizzi, S. Perrot, C. Scarpignato, T. Tolle, Towards an effective and safe [32] D. Piomelli, The endocannabinoid system: a drug discovery perspective, Curr. treatment of inflammatory pain: a delphi-guided expert consensus, in: Advances in Opin. Investig. Drugs 6 (7) (2005) 672–679. https://europepmc.org/article/med Therapy, 36, Springer Healthcare, 2019, pp. 2618–2637, https://doi.org/10.1007/ /16044662. s12325-019-01053-x. [33] D. Piomelli, A fatty gut feeling, in: Trends in Endocrinology and Metabolism, 24, [45] V. Vozella, C. Zibardi, F. Ahmed, D. Piomelli, Fast and sensitive quantification of Elsevier Current Trends,, 2013, pp. 332–341, https://doi.org/10.1016/j. δ9-tetrahydrocannabinol and its main oxidative metabolites by liquid tem.2013.03.001. chromatography/tandem mass spectrometry, Cannabis Cannabinoid Res. 4 (2) [34] K. Ren, R. Dubner, Interactions between the immune and nervous systems in pain, (2019) 110–123, https://doi.org/10.1089/can.2018.0075. ˇ ˇ ˇ in: Nature Medicine, 16, Nat Med., 2010, pp. 1267–1276, https://doi.org/ [46] S. Vuckovic, D. Srebro, K.S. Vujovic, C. Vucetic, M. Prostran, Cannabinoids and 10.1038/nm.2234. pain: new insights from old molecules, Front. Pharmacol. 9 (2018), https://doi. [35] E.M. Rock, C.L. Limebeer, L.A. Parker, Effect of cannabidiolic acid and ∆9- org/10.3389/fphar.2018.01259. tetrahydrocannabinol on carrageenan-induced hyperalgesia and edema in a rodent [47] S.J. Ward, S.D. McAllister, R. Kawamura, R. Murase, H. Neelakantan, E.A. Walker, model of inflammatory pain, Psychopharmacology 235 (11) (2018) 3259–3271, Cannabidiol inhibits paclitaxel-induced neuropathic pain through 5-HT 1A https://doi.org/10.1007/s00213-018-5034-1. receptors without diminishing nervous system function or chemotherapy efficacy, [36] E.B. Russo, Taming THC: potential cannabis synergy and phytocannabinoid- Br. J. Pharmacol. 171 (3) (2014) 636–645, https://doi.org/10.1111/bph.12439. terpenoid entourage effects, Br. J. Pharmacol. 163 (7) (2011) 1344–1364, https:// [48] A.C. Williams, B.W. Barry, Penetration enhancers, Adv. Drug Deliv. Rev. 56 (5) doi.org/10.1111/bph.2011.163. (2004) 603–618, https://doi.org/10.1016/j.addr.2003.10.025. [37] Ethan B. Russo, Beyond cannabis: plants and the endocannabinoid system, in: [49] W. Xiong, T. Cui, K. Cheng, F. Yang, S.-R. Chen, D. Willenbring, Y. Guan, H.-L. Pan, Trends in Pharmacological Sciences, 37, Elsevier Ltd., 2016, pp. 594–605, https:// K. Ren, Y. Xu, L. Zhang, Cannabinoids suppress inflammatory and neuropathic pain doi.org/10.1016/j.tips.2016.04.005. by targeting α3 glycine receptors, J. Exp. Med. 209 (6) (2012) 1121–1134, https:// [38] R. Russo, J. LoVerme, G. La Rana, G. D’Agostino, O. Sasso, A. Calignano, doi.org/10.1084/jem.20120242. D. Piomelli, Synergistic antinociception by the cannabinoid receptor agonist [50] Y.F. Zhu, K. Linher-Melville, M.J. Niazmand, M. Sharma, A. Shahid, K.L. Zhu, anandamide and the PPAR-α receptor agonist GW7647, Eur. J. Pharmacol. 566 N. Parzei, J. Sidhu, C. Haj, R. Mechoulam, G. Singh, An evaluation of the anti- (1–3) (2007) 117–119, https://doi.org/10.1016/j.ejphar.2007.03.007. hyperalgesic effects of cannabidiolic acid-methyl ester in a preclinical model of [39] P. Sacerdote, S. Franchi, S. Moretti, M. Castelli, P. Procacci, V. Magnaghi, A. peripheral neuropathic pain, Br. J. Pharmacol. 177 (12) (2020) 2712–2725, E. Panerai, Cytokine modulation is necessary for efficacious treatment of https://doi.org/10.1111/bph.14997. experimental neuropathic pain, J. Neuroimmune Pharmacol. (2013) 202–211, https://doi.org/10.1007/s11481-012-9428-2. Springer.